392
ALBERTL.HENNEAND HENRYH. CHANAN
VOl. 66
old shows the start of loss of the radiation of shorter wave length. This tendency increases with the age of the sol, until a t age eighty minutes, which is five minutes after ths gel has set, transmittance of radiation of 4000 A. is only 0.15, while that of radiation 7000 A. is 0.63 compared to a maximum of about 0.85 a t the start, as shown on the four-minute curve. These results are probably due to the preferential scattering of radiation of smaller wave length. This scattering has been studied carefully by several investigators and the results reporteda6
Summary A study has been made of the effects of radiation upon the time of set of silicic acid gels, produced by mixing solutions of sodium silicate and acetic acid. Great care was taken that no measurable temperature change was produced by the radiation. Iiadiations from a Mazda lamp, a '_---LA_ -I capillary mercury arc and an iron spark were 400 500 600 I; 10 used. Wave length in milliniicrons. No significant effects of these types of radiation Fig. 1.-Transmittance of silicic acid gel mixtures as a on the time of set were found. function of wave length and age of sol. A series of curves obtained with the Recording Spectrophotometer, showing the per cent. radiafour, forty-seven, sixty-five and eighty minutes tion transmitted by the sol are also given. The after mixing, as indicated 01: the curve. The gel of age of the sol upon light transmitted is effect set in seventy-five minutes. shown. These transmittance curves are smooth, indi( 6 ) Prasad, Mehta and Desai, J . P k y s . Ckem., 38, 1324 (1932). cating no particularly strong absorption in any N. Y. RECEIVED NOVEMBER 8. 1943 band. The curve of the mixture four minutes SCNENECTADY,
[CONTRIBUTION FROM T H E DEPARTMENT O F CHEMISTRY AT
THEOHIO
STATE UNIVERSITY]
Diolefins from Allylic Chlorides. I1 BY ALBERTL. HENNEAND HENRYH. CHANAN We have shorn' that two molecules of butadiene hydrochloride react with magnesium in ether to give a mixture of 1,5-diolefins, the composition of which remains the same whether the primary form of the halide, its secondary form, or the equilibrated mixture of both forms is used as the starting material. These results were in agreement with some of Young's observations2 on allylic equilibria. This method of synthesizing 1,5-diolefinsis now extended to five allylic chlorides grouped in a variety of pairs, Allyl and methallyl chlorides which exist in only one form (CH2=CHCH2CI and CHg=C(CHs)CH2C1) were purchased. Piperylene hydrochloride which also exists in only one form (CHsCHCICH=CHCH~) was made by adding dry hydrogen chloride t o commercial piperylene. Butadiene and isoprene hydrochlorides were made similarly from commercial (1) Henne, C h a m and 'lurk, THISJ O U H N A L , 63, 3474 (1941) (2) Young, i b i d . , 64, 404 (1932J, 68, 104 289, 4 4 1 (1938). 69, m1 2441 f1937), 60, 847, 900 (1938)
butadiene or isoprene; they both exist in two forms, namely, CHsCHClCH=CH2 S CHsCH= CHCHzCl; and (CH&C=CHCH2CI (CH& CCICH=CH2. All materials were supplied to us by a Research project of the American Petroleum Institute, directed here by Dr. C. E. Boord. Experimentally, the condensations were carried out substantially as previously' described. Two different allylic chlorides were mixed in equimolecular proportion, and the mixture was added to magnesium turnings covered with dry ether. After completing the condensation and working up the products, a hydrocarbon fraction was obtained which amounted to about 35% of the quantity possible if pure allylic halides had been used and had condensed completely. It was, however, found more expedient to use the extremely impure materials obtained from a stream of hydrogen chloride through the commercial dienes and to discard from the condensation products all impurities, unreacted halides and resinified materials. From thirty to fifty moles
393
DIOLEFINS FROM ALLYLIC CHLORIDES
March, 1944
TABLE I CONDENSATION RESULTSOF EQUIMOLBCULAR MIXTURES Allyl chloride and piperylene HCl
Methallyl chloride and piperylene
Butadiene HC1 and piperylene HC1
Allyl chloride and isoprene HC1
AA 1 4.7% MM 1 5.8% B2Bz 1 1.27i AI^ 1
4.7%
Methallyl chloride and isoprene HCI
1113
Butadiene HCI and isoprene HCl
1 4.0% B2B2 1 0.65%
B1B2 28.2 18.3%
PP 4.8 22.7% PP 6 35.4% BIBz 13.2 15.2% 1'1' 2.1 10.0% 1'1' 1.2 5.0% 1'1' 5.6 3.6% BII1 28.2 18.3yi
of each halide was used to secure a quantity of hydrocarbons ample for adequate seDaration. Before subjecting them to repeated dis'tillations through a 30-plate, helix-packed, total refluxpartial take-off, adiabatic column, the hydrocarbons were treated with sodium in liquid animoniaa until the disappearance of the Beilstein flame test indicated that they were free of halides. For physical measurements, only the centers of the distillation fractions were used. In the following list of results, the percentages of the various compounds refer only to the distilled product. Experimental Results Let each chloride be designated by its initial with a superscript (when needed) to denote a primary, secondary or tertiary isomer when two allylic isomers are possible. Thus A, M and P will designate allyl chloride, methallyl chloride and piperylene hydrochloride, while R or €3 will designate the two forms of butadiene hydrochloride, and 1' or I3 those of isoprene hydrochloride. The results of the condensations are represented in Table I, where the relative amounts of the dienes obtained by condensation are given on a molar basis. Quite a number of possible combinations were absent. In line with our previous observation' condensations between different molecules predominated over condensations between two identical molecules or two forms of the same molecule. In the fourth, fifth and sixth pairs, the ratio of the total number of I' groups to the total number of I3 groups was consistently close (3) I,ebeuu, Bid1 soc ( h i m , (8)33, 1U8Y (1905) will be published buuI1.
Total number of groups
Molecular proportions and molecular percentages of condensed dienes
Reacting chlorides
l'ractlcnl detail,
AP
A
P
15.3 72.5% MP 10 58.7% B'P 21.3 24.570
17.3
24.9
PP 22.0 25.3%
1113
AA
2.8 13.2% MI^ 2.1 8.5% 1113
8.2 5.3% B211 59.1 38.3%
2.8 13.2% MM 8.6 35.0% BIB1 11.3 7.2'&
M
P
12
22
nzP 29.3 33.8% AI1 12.5 58.9%
P 94.6
Bz 44.5
A 19.1
11
13
19.5
3.8
MI
M 31
11
13
14.9
3.1
13113
B1
13.1 8.4%
92.1
B' 89.3
1% 106.7
I' 21.3
11.7 47.5%
B1 34.5
to I 1 : P = 5:l. For the butadiene group, the ratio B1:B2 proved to be about 4:5 in the third pair, and about 1:1 in the sixth pair. In our preceding paper it had been found close to unity in the condensation of butadiene hydrochloride on itself, while it was about 5 :3 in the condensation of butadiene hydrochloride with allyl chloride. The experimental data do not permit more precise generalizations. The observed physical properties of the hydrocarbons, and the analyses of the new compounds, together with the symbols representing the diolefins appear in Table 11. Identifications Most of the diolefins were new compounds. They were e s t analyzed as reported in the table, then hydrogenated to the corresponding paraflins to establish their carbon skeleton. Most of these paraffins were known. Where the paraffins were new compounds, it proved possible to show that their physical constants were measurably different from possible alternate formulas. Finally the position of the double bonds was determined by ozone oxidation, using procedures previously described.*p5 In the following paragraphs, the observed data refer to the freezing point, the corrected boiling point a t 760 mm, the density dZo4and the refractive index n Z o D , in that order. B reduced to 4-methylheptane,8 glass, 117.6', 0.7042, 1.3980. Ozonization gave formaldehyde, acetic acid and methylsuccinic acid, m. p. 110( 4 ) Hrnne and Hill, r b l r f , 66, 761 (1!443) ( 5 ) Eieune and Perilstein, ibrd , 66, 2188 (1943) (h) Iu agreemerit with lnlorrnatrou from the Amrricau Prrrolrujll Institute.
ALBERTI,. HENNEA N I ) HBNKYH. CHANAK TABLE
Vol. 66
11
PHYSICAL CONSTANTS AND ANALYSES Name
A B
CHZ-CHCHZCHICH=CH: CH*=CHCHaCH(CHd CH=CHCHa
C
CHICH-CHCH(CHI)CH(CHJ)CH=CHCH~
D E F
CH~-C(CHI)CHZCH?C(CHI)=CHI CHnC(CHa)CHzCH(CHa)CH=CHCH;
CHi=CHCH(CHa)CH(CHdCH=CHz
c)
CHi=CHCH(CH:) CHaCH=CHCH:
H
CHr=CHCH(CHa)CH(CHi)CH=CHCHa
1
CHi=CHC( CHi)zCHzCH=CHz
CHI=CHCH~CHICH=C(CH~)CH~ K CHi-CHC(CH:)zCHzCH=C(CHdCH, L CHgC(CHs)=CHCHzCHzCH=C(CHn)CHa M CHa-C(CHa)CHzC(CHi)zCH=CHI N CHz=C(CHdCHaCHzCH=C(CHa) CH3 J
CHaCH=CHCHzCHaCH=CHCHs
P
CHn=CHCH(CHa)CHzCH=C(CHdCHa
Q
CHIC(CH~)=CHCHZCHICH=.CHCHCH~ CHa=CHC(CHi)aCHzCH=CHCHz
s
CHaCHzCHzCH(CHs)CH(CHa)CHzCHzCHa
T U
CHsCHlCH(CHa)CH(CHo)CHzCH2CHn CHaCHaC(CHa)aCHKH?CH(CHa)CHa
B. P.,
'C.
cor., "C.
Glass 64.8
-
110.3 153.3
Glass
132.1
1111
Glass Glass Glass Glass -74.4
hlI3 MI'
Glass -102 7
129.8 101.6 119.1 149.7 168.6 126.3 111 9
IIIS
0
R
Symbo1 AA AI' PP MM MP BIB' BIB? B?P AIS AI'
BIB' Bz11 B'II BLIZ
E'. p . ,
dmd Known1 0.7282 ,7662 Known1 0.7457 Known' Known: 0.7462 ,7249 ,7406 ,7657 ,7758 ,7500 ,7560 Known' 0,7491 K
Glass 134.6 N o t isolated from N o t isolated from 0 Glass 1 6 2 . 4 0.7446 Glass 140.1 7314 Glass 152 8 ,7336
IZPD
% C %H Other Calcd. Found Calcd. Found refs.
1.4231 1.4392
a.
b
b
1 4 8 1 2 87.0
87.0
1.4312 1.4160 1.4316 1.4391 1.4178 1.4288 1.4389
87.0 13.0 12.9 87.0 12.8 13.1 87.0 1 2 . 8 13.0 86.9 13.1 13.0 87.0 13.1 13.2
87.0
87.2 87.2 86.9 86.9 87.0 87.0
13.0 13 2
13.2 8 7 , l 13.0 13 1
a
c,d
86.8 13.0
e. f
1.4330 87.0 87.2 13.0 13.0 Separated as paraffin Separated as paraffin 1.4178 1.4108 1.4123
Slotterbeck, Ph.D. Dissertation, 0. S. U. (1936). * Mullikeii, Wakeman and Gerry, THISJ O U Y N A L , 57, 1605 (1935). Kresterisky, J. Russ. Phys.-Chem. Soc., 46,904 (1914). Petrov, J . Gen. Chem. (U.S. S. R.),9, 2129 (1939). e Harries and Weil, Ber., 37, 846 (1904). Escourrou, Bztll. soc. chirn., [ 4 ] 39, 1249 (1926).
'
111'. C reduced to a new parafin (S in table) measurably different froni the possible known isomers. Ozonization gave acetaldehyde, acetic acid and a dicarboxylic acid m. p. 1'78-190' with a molecular weight of 147.4. This acid was regarded as a mixture of dl- and cr,&dimethylsuccinic acid. The pure meso-acid is said to melt7 at 198' and 208' with decomposition, while the dl-mixture melts' at 12T'; they can be separated by treatment with a benzene-ether mixture, in which the dl-form is more soluble than the nzesoform. This treatment was applied to the dicarboxylic acid, and raised the ni. p. to 18s-189.5'. Further purification was deemed not to be indispensable, in view of the fact that the isomeric diacids melt much lower, i. e., ethylsuccinic acid at loo', and a,a-dimethylsuccinic at 141°. E reduced to 2,4-dimethylheptane,' glass, 13S.1°, 0.7153, 1.4031. Ozonization gave formaldehyde, acetic acid and a keto-acid whose semicarbazone had m. p. 178" and analyzed correctly for C7H13N303 (calcd.C, 44.9; H, 7.0; r\i, 22.4. Found: C,44.9; H, 7.0; N, 22.3). The keto acid was interpreted as being a-methyl-7-ketovaleric acid, whose semicarbazone is said to meltYat 191' and also a t 181'. H reduced to u new paraffin (T in table) measurably cliff erent from the known isomers. Ozonization gave foriiddehyde, acetic acid and C Y , @-dimethylsuccinic acid in. p. 190' (cf. compound C) I reduced to 3,3-dirnethylhexane,G glass, ll?..L'", 0.7096, 1.4000. Ozonization gave formaldehyde and a,a-dimethylsuccinic acid m. p. 139". J reduced to 2-meth~lheptane,~ - 10'3.6', l17,(j0, 0.6978, 1.3950. Ozonization gave formaldehyde, acetone with a 2,4-dinitrophenylhydrazone :7) Heilbron, "Dictionary of Organic Compoiinds." 8) Richards and Shipley. ?'HIS J ~ I . R x . % I . ,38 i.9) BCbsl. Buil. JOG. chii,i., 1 25, ?i.; '1'1,J: ,ilia 8 ,
I
w,ft.r
I
t-/t,tt,
1212 ' I
R3ki.r- . i i i c l
m. p. 185' and succinic acid m. p. 186'. K reduced to a new paraffin (U in table) measurably different from the known possible isomers. Ozonization gave formaldehyde and a,a-dimethylsuccinic acid m. p. 141'. L reduced to 2,7160.0' ; 0.7250; dimethyloctane,lo - 50.7' ; 1,40S7. Ozonization gave acetone detected b the nitroprusside test, and succinic acid m. 187Iy. M reduced to 2,4,4-trimethythe~ane,~~ - 117.7', 130.8', 0.7231, 1.4068. Ozonization gave formaldehyde and a,a-dimethyl- y-ketovaleric acid, with a semicarbazone, m. p. 196'. N reduced to 2,6dimethylheptane,12 - 103.6', 135.2', 0.7102, 1.4009. Ozonization gave formaldehyde, acetone (color test) and levulinic acid with a semicarbazone m. p. 184.5' (calcd., N, 24.2. Found: N , 23.9). P reduced to 2.5-dimethylheptane,13 glass, 1:36.2', 0.7175, 1.4042. Ozonization gave formaldehyde, acetone (color test) and methylsuccinic acid, m. 110'. Q could not be isolated from K. The mixture was hydrogenated and the parafins were separated. The paraffin ratio was taken to represent the diolefin ratio in the mixture. R could not be isolated from 0. This ratio was computed from the relative quantities o f paraffins, after hydrogenation. Summary Five allylic chlorides have been condensed in six different pairs t u give a series of l,5-diolefins, which were purified sufficiently to permit the listing of their physical properties and the determination of their relative abundance. CULCMBU O~m, ) -
RECEIVED OCTOBER 20, 1943
(1936).
101 Calingarrt Tais J O U R N A L , 68, 636 1I ) \lo,her. Ph D Dissertation, Pennsylvania State College. 1'440 12) U hite, J Rcidurch VofL Bur. Slondards, I), 317 (19101 1 7 ci A r l ' ~ ueKg9 IW. i o r I r v * r 14, 54 f i w l )